The HIV epidemic presents urgent challenges to develop better methods to promote the adoption and maintenance of sexual behavior changes to reduce risk for contracting HIV infection. Conceptual models must be refined to develop more effective approaches for face-to-face behavior change interventions as well as for large-scale community-level or population- based behavior change trials. Outcome evaluation of group and community- level HIV prevention interventions is the focus of research among many CAIR investigators and we have a strong reputation for conducting high- quality HIV prevention outcome research. Core A, Intervention Model Development, will support this area of research by: (1)Sponsorship of annual behavior change intervention models development working groups which will bring together major scientists in this theoretical area; (2) planning internal and cross-project analyses of existing CAIR intervention outcome data sets to evaluate intervention potency and generality across multiple projects; (3) Establishment of a core function to monitor enrollment and retention of ethnic minorities and women across studies, and identify research protocol characteristic related to successful or unsuccessful recruitment and retention; (4) Development of assessment and sampling methods to study social, psychological, service availability, and other characteristics of communities that predict HIV risk levels and response to HIV prevention interventions; and (5) Establishment of an intervention integrity assurance component that will systematically review records (e.g., audiotapes or videotapes) of intervention sessions in study protocols to determine correspondence of interventions delivered with conceptual components underlying independent variable manipulations in these studies. These functions will benefit existing, new, and developmental studies by CAIR investigators, and will facilitate basic research in intervention model development.